Python: [Bug]: A2AAgent should not ignore _update_event when processing responses

[lixiaoqiang]

Description

The current implementation ignores the _update_event variable, which causes the streaming output of the connected agent to fail to obtain typewriter-effect output content.

        async for item in a2a_stream:
            if isinstance(item, A2AMessage):
                # Process A2A Message
                contents = self._parse_contents_from_a2a(item.parts)
                update = AgentResponseUpdate(
                    contents=contents,
                    role="assistant" if item.role == A2ARole.agent else "user",
                    response_id=str(getattr(item, "message_id", str(ulid.ULID()))),
                    raw_representation=item,
                )
                all_updates.append(update)
                yield update
            elif isinstance(item, tuple) and len(item) == 2 and isinstance(item[0], Task):
                task, _update_event = item # ~~~JUST HERE !!!! ~~~~
                for update in self._updates_from_task(task, background=background):
                    all_updates.append(update)
                    yield update
            else:
                raise NotImplementedError("Only Message and Task responses are supported")

The streaming response workflow of the A2A Server is roughly as follows:

1. The first event returns a Task
2. The second event returns the first chunk of the model output wrapped as a TaskArtifactUpdateEvent, with append = False and TaskState.working

		TaskArtifactUpdateEvent(
                    task_id=task_id,
                    context_id=context_id,
                    artifact= The First Chunk Of Model Output,
                    append= False,
                )

3. Subsequent events wrap the subsequent model outputs as TaskArtifactUpdateEvent in sequence, with append = True and TaskState.working
4. Returns a TaskStatusUpdateEvent with TaskState.completed

To be honest, after using the A2AAgent implementation for a period of time, I have encountered many issues, submitted some of them, and they have been fixed. First of all, I would like to thank all the developers for your hard work and timely responses.

However, this also exposes a problem: the implementation of A2AAgent has not been properly integration-tested in real business scenarios. It is already at the RC5 version, and the current quality is worrying.

Code Sample

Error Messages / Stack Traces

Package Versions

agent-framework:1.0.0rc5

Python Version

No response

Additional Context

No response